IEEE Computational Intelligence Magazine - May 2021 - 62

Research
Frontier

Manuel Chica
University of Granada, SPAIN
Juan M. Hernández
University of Las Palmas de Gran Canaria, SPAIN
Casiano Manrique-de-Lara-Peñate
University of Las Palmas de Gran Canaria, SPAIN
Raymond Chiong
The University of Newcastle, AUSTRALIA

An Evolutionary Game Model
for Understanding Fraud in Consumption Taxes
Abstract

his paper presents a computational evolutionary
game model to study and
understand fraud dynamics in
the consumption tax system.
Players are cooperators if they
correctly declare their value
added tax (VAT), and are defectors otherwise. Each player's payoff is influenced by the amount
evaded and the subjective probability of being inspected by tax
authorities. Since transactions between
companies must be declared by both
the buyer and seller, a strategy adopted
by one influences the other's payoff.
We study the model with a wellmixed population and different scalefree networks. Model parameters were
calibrated using real-world data of
VAT declarations by businesses registered in the Canary Islands region of
Spain. We analyzed several scenarios of
audit probabilities for high and low
transactions and their prevalence in
the population, as well as social
rewards and penalties to find the most
efficient policy to increase the proportion of cooperator s. Two major
insights were found. First, increasing
the subjective audit probability for low
transactions is more efficient than

Digital Object Identifier 10.1109/MCI.2021.3061878
Date of current version: 12 April 2021

duction minus value of intermediate consumption) generated at
each step of the production and
distribution process. The tax debt
is calculated by applying the tax
rate to the value of the goods and
services sold, and deducting the
VAT attached to intermediate
consumption. Companies and
professionals declare the VAT
passed on to the customers, while
©SHUTTERSTOCK.COM/LEREMY
deducting the amount borne
from purchases from their own
increasing this probability for high
suppliers. The resulting statement can be
transactions. Second, favoring social
either positive or negative.
rewards for cooperators or alternative
This refunding system potentially
penalties for defectors can be effective
allows for significant levels of fraud in
policies, but their success depends on
terms of undervaluation of sales and
the distribution of the audit probabiliovervaluation of purchases [2]. However,
ty for low and high transactions.
the fact that both buyers and sellers
record each transaction offers some selfenforcement capability to the VAT sysI. Introduction
tem. Das-Gupta and Gang [3] identified
The value added tax (VAT) is the most
circumstances where the ability of tax
common consumption tax worldwide.
administrators to match the sales and
With extensive use since the 1960s, it
purchase invoices strengthens this selfhas reached significant tax revenue
reinforcement capacity. In any case,
capacity. In fact, consumption taxes
tracking the relationships between buydemonstrated a similar collection capacers and sellers is clearly of the utmost
ity to income taxes during the 1990importance to detect and prevent
2010 period [1]. All deliveries of goods
fraud. For this purpose, tax administraand services by companies, professionals,
tion normally requires the declaration
and importers are subject to VAT. The
of all purchase and sales transactions
tax base is the value added (value of probetween pairs that exceed a certain
threshold, with an explicit declaration
Corresponding Author: Manuel Chica (Email: manuelchica@
of the counterparts.
ugr.es).

IEEE COMPUTATIONAL INTELLIGENCE MAGAZINE | MAY 2021

http://www.SHUTTERSTOCK.COM/LEREMY

IEEE Computational Intelligence Magazine - May 2021

Table of Contents for the Digital Edition of IEEE Computational Intelligence Magazine - May 2021